PROJECT SUMMARY Acute lymphoblastic leukemia (ALL) is the most prevalent childhood cancer, and although overall survival rates of ALL have substantially improved, resistance to antileukemic agents remains a major clinical problem. Antileukemic drug resistance is predictive of poor disease outcome and is commonly observed in ALL patients that have relapsed, who have a low overall survival rate of only 40%. The mechanisms that cause ALL relapse and drug resistance remain poorly understood. To address how inherited genomic variability contributes to these mechanisms, genome wide association studies (GWASs) have identified DNA sequence variation associated with ALL treatment outcome. However, since these variants are noncoding in nature, their connection to gene function, ALL biology and antileukemic drug resistance has been difficult to establish. Moreover, given that hundreds of variants are typically in strong linkage disequilibrium with the associated variant, pinpointing causal variants at GWAS loci has been challenging. To address these challenges, we have generated functional genomic maps of the ALL genome, including the precise locations of noncoding regulatory elements in >40 ALL samples. Through rigorous open chromatin-fine mapping using our ALL genome maps, and by integrating our results with drug resistance phenotypes from primary ALL cells obtained from patients at St. Jude (SJ), we identified 3229 variants at 125 GWAS loci associated with ALL treatment outcome that are predicted to have an impact on gene regulation, leukemic cell biology and antileukemic drug resistance. Using these fine-mapped variants, we propose an integrative strategy for identifying candidate causal variants associated with ALL treatment outcome, and a rational experimental system for functionally linking these variants and their target genes to antileukemic drug resistance. In Aim 1, we will perform massively parallel reporter assays (MPRAs) on >3200 fine-mapped variants to assess their gene regulatory activities and to identify allele-specific differences in activity. In Aim 2, we will employ a polygenomic strategy by integrating our MPRA results with diverse genomic datasets to prioritize and rank fine-mapped variants by their likelihood of being causal. For this effort, we will capitalize on the unique and rich resources available at SJ, including extensive genomic characterizations and drug resistance phenotypes from large ALL patient cohorts, as well as ongoing genomic characterizations for all new and/or relapsed patients. We will functionally validate the role of the top 20 ranked candidate causal variants on antileukemic drug resistance using CRISPR technology and chemotherapeutic drug viability assays in human ALL cell lines. We will also identify GWAS target genes and functionally assess their role in antileukemic drug resistance in human ALL cell lines and in patient-derived xenograft mouse models of pediatric ALL. Collectively, our proposal will uncover novel genomic mechanisms of antileukemic drug resistance. Ultimately, these data can inform approaches to circumvent resistance in the clinic, and be used to improve initial treatment, as well as guide therapy for relapsed disease through precision medicine and more personalized treatment regimens.